Sociogenomics

Fashionable pseudoscience

Georgy Golovanov, Hi-tech+

Relatively easy access to genetic information and the development of biotechnologies have given rise to sociogenomics, a discipline that tries to describe human behavior based on his genome. But, even if we leave aside the ethical side of the issue, experts have doubts about the scientific nature of this direction.

Since the discovery of the DNA double helix, there have been several scientific "revolutions", and now we are in the middle of another one of them. We are talking about the union of social and natural sciences with big genome data, which, thanks to companies like 23andMe, attribute genetic justifications to human behavior.

These studies are conducted by young, often charismatic scientists who write popular science books, willingly give lectures and give interviews. Their activity proves that the discussion about nature and upbringing has not stopped, it has just moved to a new level, writes MIT Technology Review.

Defenders of sociogenomics see prospects that not everyone will consider a boon. For example, health reports based on genetics and issued at birth predict the risk of diseases and a tendency to certain behaviors.

Sociologists study the genetic predisposition to academic success and high position in society, and economists study the risk of issuing a loan based on an inherited propensity to spend or hoarding.

Sociogeneticists usually talk about the attractive side of these changes. In addition, they believe that nothing can be changed. "The gin broke out of the bottle," says psychologist Robert Plomin, "and you can't drive it back."

Skeptics point out that it is incorrect to call sociogenomics a science so far. Her conclusions are not confirmed experimentally and do not proceed from general principles or hypotheses. Also, it does not have a foundation in the form of a multi-year collection of evidence, such as, for example, geology or biology. And specialists act in an uncharacteristic way for other sciences: they usually collect data first, and then a hypothesis arises.

The data collected due to the widespread use of genetic tests are loaded into algorithms that produce correlations between the trait of interest and a tiny DNA sequence the size of one nucleotide. Then sociogeneticists deduce patterns from this information and make predictions, mainly about the behavior of an individual in the future.

Bob: Scientist or fool

Biologist Graham Coop proves that big data can lead to a dead end. As an example, he conducts a thought experiment. Suppose the purpose of the study is to answer the question, why do the British drink more tea than the French?

Imaginary researcher Bob uses data from a genetic database. He counts the average number of alleles (different forms of the gene) associated with tea preference among the British and the French. If the former have more of them, then Bob can claim that the British love of tea is explained by genetics.

However, the problem is that the same alleles may behave differently in other genomes and under different circumstances. In other words, we cannot be sure that another group of people with the same genetic variants will also prefer tea. And even if it is, we still don't know if it was caused by genes.

Bob makes two mistakes: first, he confuses correlation with causality. The study does not show that alleles associated with tea drinking affect tea consumption. Secondly, statistical significance is not equal to biological significance. The number of people buying ice cream on the beach correlates with the number of drowned or eaten by sharks.

Ice cream sales can predict the number of victims, but only a fool will knock a cup out of your hands in full confidence that he saved you from death.

"Complex traits are really complex," says Coop. "Most traits are incredibly polygenic and include tens of thousands of loci. These loci are like a huge number of pathways associated with many environmental and cultural factors."

Portal "Eternal youth" http://vechnayamolodost.ru